Performance evaluation of a tandem queueing network Smail Adjabi 1 and Karima Lagha 2 1 Laboratory LAMOS University of Bejaia 06000 Bejaia, Algeria (e-mail: adjabi@hotmail.com) 2 Laboratory LAMOS University of Bejaia 06000 Bejaia, Algeria (e-mail: lagha@yahoo.fr) Abstract. In order to evaluate the characteristics of a tandem queueing network, we propose a study, taking into account the qualitative properties of distributions. For this, we consider different bounds (lower and upper bounds) for different classes of nonparametric distributions. These bounds are computed while applying the QNA method (Queuing Network Analyser). To verify whether the proposed inter- vals include(contain) the approximate values, we have considered some approxima- tions as those corresponding to KLB (Kramer Langenbach Benz) and simulation methods. Two algorithms have been constructed for programming the methods, and implemented for under the assumption that the inter-arrival distribution of the network is parametric or nonparametric. Keywords: Queueing networks, nonparametric distribution, simulation. Introduction An efficient designing of networks cannot be performed without knowing their performances. Performance evaluation allows, for instance, to: - compare different topologies and protocols of networks according to their application and the service offered. - anticipate and correct eventual problems related to performances. There are three techniques for performance evaluation: Technique of Data (direct measurement), analytical technique and the simulation technique. In the present work, we consider a queueing network in tandem. A perfor- mance evaluation is carried out by using the qualitative property of distri- butions for determining a lower and upper bound for the characteristics of the network. These bounds will be computed, after determining the para- meters of the distributions characterizing the inner flows by using the QNA method(Queueing Network Analyzer). In view to verify whether the proposed intervals contain the approximate values, we have applied the KLB(Kramer and Langenbach-Belz) and the simulation methods. 1 The model We consider the following queuing network: